developing a benchmarking model for construction...

Journal of Construction Engineering and Management (JCEM), ASCE

(Final Accepted Manuscript), Volume 139, Issue 6, June 2013, Pages 705-716

1

Developing a Benchmarking Model for Construction

Projects in Hong Kong

John F.Y. Yeung1; Albert P.C. Chan2; Daniel W.M. Chan, M.ASCE3;

Y.H. Chiang4; and Huan Yang5

1 Lecturer I, College of International Education, School of Continuing Education,

Hong Kong Baptist University, Hong Kong; formerly Postdoctoral Fellow,

Department of Building and Real Estate, The Hong Kong Polytechnic University,

Hung Hom, Kowloon, Hong Kong, China (corresponding author). E-mail

address: [email protected]

2 Professor and Associate Dean, Faculty of Construction and Environment, The

Hong Kong Polytechnic University, Hung Hom, Kowloon, Hong Kong, China.

3 Associate Professor, Department of Building and Real Estate, The Hong Kong

Polytechnic University, Hung Hom, Kowloon, Hong Kong, China.

4 Professor, Department of Building and Real Estate, The Hong Kong

Polytechnic University, Hung Hom, Kowloon, Hong Kong, China.

5 Lecturer, Institute of Finance and Economics Research, Shanghai University of

Finance and Economics, Shanghai, China.

Abstract: This paper attempts to incorporate both leading and lagging KPIs and

apply the reliability interval method (RIM) to formulate a benchmarking model

to assess project success in Hong Kong. A list of leading and lagging KPIs was

complied based on a comprehensive literature review. This list of KPIs was used

to develop a survey questionnaire and RIM was subsequently used to analyze

the survey results and determine the relative importance and rankings of

various leading and lagging KPIs. The results reveal that the top 10 KPIs to

evaluate the success of construction projects in Hong Kong (in descending order)

were: (1) safety performance; (2) cost performance; (3) time performance; (4)

quality performance; (5) client’s satisfaction; (6) effectiveness of communication;

(7) end user’s satisfaction; (8) effectiveness of planning; (9) functionality; and (10)

environmental performance. Finally, a composite performance index (CPI) was

derived by means of RIM to provide a comprehensive assessment of

construction project success in Hong Kong. Different construction projects can

now be assessed on the same basis for benchmarking and project monitoring

This is the Pre-Published Version.



2

purposes. Construction senior executives and project managers can thus use the

CPI to measure, evaluate and improve the performance of their construction

projects at various stages of the project life cycle (i.e. pre-planning phase,

planning phase, design phase, construction phase, and commissioning phase).

Although the CPI was developed locally in Hong Kong, the research methods

could be replicated elsewhere to produce similar indices for international

comparisons. Such an extension would enhance the understanding of managing

construction projects across different places.

CE Database subject headings: Benchmark; Construction management;

Hong Kong.

Author keywords: Benchmarking; Key performance indicators (KPIs);

Composite performance index (CPI); Reliability interval

method (RIM); Hong Kong.

Introduction

Over the past decade, many researchers have applied the concept of Key

Performance Indicators (KPIs) to conduct benchmarking studies in the

construction management discipline (Cox et al. 2003; Chan and Chan 2004; Lee

et al. 2005; Costa et al. 2006; Yeung et al. 2007, 2009). However, Beatham et al.

(2004) and Costa et al. (2006) criticized that the most significant problem with

the KPIs in their current format is that they do not offer the opportunity for

organizational change. They are designed to be used as ex-post “lagging”

KPIs, which are used to assess performance results of completed projects.

They do not offer the opportunity to change performance or alter the result of

associated performance of on-going projects. They are used only as a historic

review. Therefore, the use of leading KPIs was advocated to provide early

warnings, identify potential problems, and highlight any needs for further

investigation or actions. Leading KPIs do offer the opportunity to change.

They are measures of performance whose results are used to predict future

performance of the activity being measured and present the opportunity to

change practice accordingly, or to enable future decisions to be made on future

associated activities based on the outcome of previous activities. In fact, few, if

any, researchers have tried to incorporate both leading and lagging KPIs to

compile a Composite Performance Index (CPI) for measuring the overall

performance of construction projects, thus making benchmarking and project

monitoring ineffective. In addition, they seldom consider the impact of the

fuzziness of an individual KPI on the evaluation of the performance of a



3

construction project. This paper, therefore, aims to incorporate both leading

and lagging KPIs and apply RIM to compile a CPI to assess the performance of

construction projects in Hong Kong. By doing so, different construction

projects can be evaluated on the same basis for benchmarking and project

monitoring purposes. Construction senior practitioners and project managers

can thus adopt the CPI to measure, evaluate and improve the performance of

their construction projects to achieve excellence at different stages of the project

life cycle.

Background of Study

Many researchers have conducted research on project performance evaluation

and benchmarking in the construction management discipline. Cox et al. (2003)

stated that there is a great need in the construction industry for identifying a set

of common indicators to be used by construction executives and project

managers in measuring construction performance at the project level. The

focus of the research was to collect management perceptions of the KPIs

currently utilized in the construction industry. They concluded that six

indicators, including (1) quality control; (2) on-time completion; (3) cost; (4)

safety; (5) $/unit; and (6) units/MHR were reported as being most useful by

every segment of the construction industry. Chan and Chan (2004) developed

a framework for measuring success of construction projects. The framework is

composed of a set of KPIs, which are measured both objectively and subjectively.

The validity of the proposed KPIs was also tested by three case studies. Lam et

al. (2007) developed a project success index (PSI) to benchmark the performance

of design-build projects from a number of KPIs. The findings showed that

time, cost, quality, and functionality should be the principal success criteria for

D&B projects. Yeung et al. (2007) developed a model using the Delphi survey

technique to objectively measure the performance of partnering projects in

Hong Kong. The results indicated that the top-7 weighted KPIs to evaluate the

success of partnering projects in Hong Kong were: (1) time performance; (2) cost

performance; (3) top management commitment; (4) trust and respect; (5) quality

performance; (6) effective communications; and (7) innovation and

improvement. Finally, a composite Partnering Performance Index (PPI) for

partnering projects in Hong Kong was derived to provide an all-round

assessment of partnering performance. Later, Yeung et al. (2009) applied the

same Delphi survey technique to formulate a model to assess the success of

relationship-based construction projects in Australia. The Delphi survey

selected eight KPIs to evaluate the success of relationship-based projects in

Australia. These KPIs included: (1) client’s satisfaction; (2) cost performance;

(3) quality performance; (4) time performance; (5) effective communication; (6)



4

safety performance; (7) trust and respect; and (8) innovation and improvement.

An equation for calculating a performance index for relationship-based projects

in Australia has been finally derived. It should be highlighted that these

developed indices were composed of a set of lagging KPIs and they could be

used to measure, monitor, and improve the performance of their completed

construction partnering and relationship-based projects. This paper provides

an objective basis for measuring and predicting project performance at different

stages of the project life cycle by means of both leading and lagging KPIs.

Research Methodology

The research methods employed included: (1) a comprehensive and critical

literature review; (2) empirical questionnaire survey; and (3) weighting

assessment relating to the empirical questionnaire survey results by means of

RIM.

Literature Review

In order to acquire a clear understanding of benchmarking related research, this

study carried out a two-stage literature review to conduct a content analysis of

benchmarking related articles from 1998 to April of 2009, which is presented in

Figure 1.

In stage 1, the search engine “Scopus” was used to conduct the literature search

under the “Title/Abstract/Keyword” field. Search keywords included “Key

Performance Indicators”, “KPIs”, “Benchmark”, “Benchmarking”, “Project

Success”, “Performance Measurement”, “Performance Measure”, “Critical

Success Factors”, “CSFs”, “Critical Success Factor”, “CSF”, “Best Practice”,

“Best Practices”, and “Continuous Improvement”. Over thirty thousand

(30,766) articles were so identified on 14 April 2009. The search included many

irrelevant publications. To narrow down the search, only the construction

journals that have published the most benchmarking papers were chosen. The

results in stage 1 revealed that International Journal of Project Management,

Construction Management and Economics, Journal of Construction Engineering and

Management, Journal of Management in Engineering, and Engineering, Construction

and Architectural Management have published the most benchmarking related

papers in construction, and were thus selected as target journals in stage 2. As

Benchmarking: An International Journal has published the largest number of

benchmarking related studies, of which most are related to construction, it is

also included as a target journal in stage 2. It should be noted that these

journals were also ranked high in the ranking of Chau (1997). These selection



5

processes reinforce each other. The number of papers so identified was

trimmed down to 572.

Figure 1. Two-stage literature review for this Study (Ke et al. 2009)

Table 1 shows the level of relevance of the 572 papers with benchmarking in

construction, as evaluated according to their abstracts. It shows that 14 papers

are classified as “Very Relevant” or above on benchmarking in construction

(Table 1). These 14 papers were fully reviewed and a total of 56 performance

indicators (including both leading and lagging performance indicators) were

Search engine

T/A/K search

Papers

Number per year Journals

Chau’s ranking

Stage 1

Stage 2 Target journals

Visual Examination

Affiliation

Institutions Countries

Papers

Interests Citation

T/A/K search

Researchers

Validate

Note: T/A/K - Title/Abstract/Keywords



6

identified (Table 2). Since the meanings of some performance indicators are

similar in nature, they were combined together. For example, the meanings of

“cost performance”, “project cost growth (owner)”, “project budget factor

(contractor)”, “change cost factor”, “predictability cost”, “factor phase cost

growth (owner data only)”, and “phase cost factor (owner)” are similar in

nature, they are combined together as “cost performance”. Other performance

indicators, such as time performance, safety performance, and quality

performance, are combined in a similar way. By using this method, a total of

28 performance indicators were thus derived (Table 3). Furthermore, 8 of them

were cited for one time only, they were regarded as less important so they are

not chosen for further analysis. Finally, a total of 20 KPIs were selected for the

development of the survey questionnaire. It should be noted that 7 out of the

20 KPIs, including (1) productivity performance; (2) project team satisfaction; (3)

effectiveness of risk management; (4) effectiveness of planning; (5) provision of

training courses; (6) effectiveness of material management and resource

management; and (7) effectiveness of communication, are leading measures

because they offer the opportunity to change performance or alter the result of

associated performance of on-going construction projects.

Questionnaire Survey

The questionnaire was then developed based on the KPIs identified from the

comprehensive and critical literature review. The results of the survey would

be used to develop appropriate weightings for different KPIs. There were two

major sections in the questionnaire, encompassing: (1) background information

of survey respondents; and (2) evaluation of appropriate weightings for a list of

KPIs to assess the project success in Hong Kong. These two parts were to be

answered by survey respondents with interval grading so that RIM can be

facilitated. A total of 1,200 self-administered blank questionnaires were sent to

target industrial practitioners via postal mail and 233 completed questionnaires

were returned, representing an acceptable response rate of 19.42%. The

professional affiliation for the 233 survey respondents included architect (7.7%),

building surveyor (6.4%), quantity surveyor (7.3%), project manager (10.7%),

engineer (51.1%), builder (4.7%), and others (12%). Most of them worked in

either client organizations or main contractors (Table 4). A majority of them had

more than 10 years of professional working experience (Table 5). It is believed

that they possessed adequate knowledge and rich hands-on experience to

evaluate the importance of each KPI on construction projects in Hong Kong.

In order to assess the weighting for each KPI, the proper weighting assessment

method must take into account the fuzzy nature of its perceived importance.

The following section presents two possible weighting assessment methods.



7

Weighting Assessment Methods

It is important to choose the appropriate weighting assessment method since

this directly affects the accuracy of evaluating the success of a construction

project. Two weighting assessment methods were considered in this study.

They are: (1) the Analytical Hierarchy Process (AHP); and (2) RIM.

AHP is a theory of measurement adopting pair-wise comparisons. It relies

much on the judgments of experts to derive priority scales. It is these scales

that measure the intangibles in relative terms. The comparisons are made

using a scale of absolute judgments that represent the extent to which one

element dominates another with respect to a given attribute (Saaty 2008). Yiu et

al. (2005) reported that there are two major limitations of AHP. First, it is

difficult to avoid inconsistency between pair-wise comparisons even if assessors

have comprehensive explanations of the factors and sub-factors. Second,

assessors may find it difficult to determine an exact weighting for some factors

because they are vague in nature. In other words, the use of AHP is not able to

cope with fuzziness satisfactorily and assessors may not be able to provide

appropriate weightings when they find it difficult to weigh these vague criteria.

These limitations make AHP not suitable in this study.

Reliability Interval Method (RIM)

With reference to Moore’s (1979) research work, Lo et al. (2001) developed RIM

to assess fire risk for high-rise buildings. RIM is particularly useful in

handling imprecise information. It requires assessors to weigh a factor using a

fuzzy range of numbers. For example, assessors can weigh a factor as a range

of 3 to 5, [3, 5], instead of an exact value of [4]. The influence of a KPI on the

performance of a project is greater if the weighting is higher. Since pair-wise

comparisons are not needed in this assessment method, the problem of

inconsistency arising from pair-wise comparison is removed. This method can

also determine the degree of reliability, center variance (CV), and interval

variance (IV). According to Lo et al. (2005), the degree of reliability is the

proportion of the ranges weighted by the assessors which falls within the

average range. CV and IV indicate the consistency of opinions amongst survey

respondents. Yiu et al. (2005) used RIM to develop weightings for different

decision criteria and their sub-criteria in evaluating cost estimator’s

performance. Lo et al. (2005) stated that this method is particularly practical

when the number of factors and sub-factors are large because the use of

pair-wise comparisons in AHP may lead to a lengthy questionnaire. RIM was

therefore chosen for this research as an appropriate weighting assessment

method for determining the weightings of each KPI. This method mitigates the



8

problem of fuzziness during the evaluation of appropriate weightings for a list

of KPIs.

Two assumptions are made for the development of RIM:

(1) The probability distribution function within the interval is linear;

(2) Equal weighting is given to the opinions of the different experts.

To allow statistical analysis of the results, RIM provides three parameters with

the fuzzy assessment of weightings, namely, reliability, center variance (CV) and

interval variance (IV). The reliability illustrates the proportion of the ranges

weighted by the experts that falls within the average range. The value of CV

reflects the difference between the grade eigenvalue ( jζ ) and the average of

interval grades (lij and rij) for a particular KPI. The value of IV reflects the

difference between the grade eigenvalue ( jζ ) and the interval value lij or rij

(which has a larger difference with the grade eigenvalue). The consistency of

opinions among experts can be reflected with the use of these two variances.

The smaller the values of center variance (CV) and interval variance (IV) are, the

more consistent are the opinions of the respondents (Lo et al. 2001, 2005; Yiu et

al. 2005).

Research Findings and Discussions

Table 6 indicates the results of the average interval grades for each KPI obtained

from the 233 completed and valid questionnaires. Table 7 shows the survey

results of the respondents’ weightings of each KPI. Since RIM has only been

recently applied to the construction management discipline, the following

principles should be noted. Lo et al. (2001) stated that when the weightings of

two factors are nearly the same, the one with the higher reliability and the vice

versa is actually more reliable. They further stated that a low value of variance

indicates that a higher level of consistency exists amongst survey respondents,

and vice versa. Yiu et al. (2005) conducted questionnaires on performance

evaluation for cost estimators and suggested that when adopting RIM, an

achievement of 65% in reliability could be regarded as “reasonably good”.

They also took the view that only minor inconsistencies in opinions exist

amongst clients if the values of average center and interval variances are lower

than 0.65 and 2.10 respectively. Accordingly, cut-off values of 0.65, 0.65 and

2.10 for reliability, center variance (CV) and interval variance (IV) respectively

were used in selecting appropriate KPIs in this study. In total, 18 KPIs met

these requirements and the top-10 weighted KPIs were selected for compiling

the CPI to assess the success of a construction project in Hong Kong. The



9

top-10 weighted KPIs were selected because all of their weightings are larger

than 0.05. The rationale for choosing 0.05 as the cut-off value is mainly

because 0.05 is the average value of all the KPIs’ weightings. Since the

weightings of top-10 weighted KPIs are larger than 0.05, they were selected to

form the CPI.

Table 7 portrays the results of rankings and weightings of the 20 identified KPIs

in descending order. The results reflect that “Safety Performance” is

considered as the most important KPI for construction projects in Hong Kong.

Its reliability nearly reaches 82%, which is satisfactory. The center variance

(CV) and interval variance (IV) of this KPI are small (0.32490 and 0.85009

respectively) and this implies that the opinions of the survey respondents are

very consistent. With high reliability and small variances, the top rank of this

KPI is justified. In fact, “Safety” becomes the most core value in many

contracting firms. Zero Harm Target was advocated by a leading contractor in

Hong Kong since 2010 (Gammon Construction 2010). PASS scores administered

by the Hong Kong Housing Authority (HKHA) are composed of a major

component on safety issues. If a contractor scores low in PASS (Performance

Assessment Scoring System), its tendering right might be jeopardized which is

an effective means to ensure safety compliance from the contractor (Chan et al.

2006). Similar policies apply in other public works projects administered by

the Hong Kong SAR Government. If a contractor achieves a poor safety record,

its senior management needs to be interviewed by the Contractors Registration

Committee (CRC) under the Buildings Ordinance before its registration can be

renewed. The second and the third most important KPIs are “Cost

Performance” and “Time Performance”, with weightings of 0.05763 and 0.05762

respectively. These two KPIs achieve a reliability of nearly 93% and 95%

respectively and their CV (0.28111 and 0.24727, respectively) and IV (0.93499

and 0.91820, respectively) are also small, which show that the opinions of

survey respondents are consistent.

It is commonly accepted that project success is measured by project

performance in terms of cost, time, and quality (Chan et al. 2002). Our

research findings are consistent with this traditional wisdom because cost

performance, time performance, and quality performance take the second, third,

and fourth positions respectively in this study. In addition, the findings also

emphasize the importance of safety performance and client’s satisfaction. This

implies that industrial practitioners in Hong Kong take safety seriously. In fact,

the accident rate per 1000 workers has decreased from 247.9 in 1998 to 60.0 in

2007 (Hong Kong Labor Department 2009). In order to compile a composite

indicator to evaluate the performance of construction projects in Hong Kong, a



10

Composite Performance Index (CPI) is developed which can be represented by

the following formula:

CPI = 0.1086 x safety performance + 0.1058 x cost performance

+ 0.1058 x time performance + 0.1018 x quality performance

+ 0.0991 x client’s satisfaction + 0.0969 x effectiveness of communication

+ 0.0968 x end user’s satisfaction + 0.0959 x effectiveness of planning

+ 0.0955 x functionality + 0.0937 x environmental performance

Note: Italic fonts refer to “leading” KPIs

The coefficients are the “normalized” individual weightings of the top-10

weighted KPIs, which are calculated by dividing their individual weightings by

the total weightings of top-10 weighted KPIs. For example, the coefficient of

safety performance is calculated by 0.05914 ÷ (0.05914 + 0.05763 + 0.05762 +

0.05541 + 0.05396 + 0.05274 + 0.05270 + 0.05226 + 0.05202 + 0.05100), which is

equal to 0.1086. The CPI is derived based on the assumption that this is a

linear and additive model (Frisch et al. 1992; Curtin 1982; Nielsen 2001;

Ramaswami and Roe 2004; Yeung et al. 2007 & 2009; Xu et al. 2010). It is

logical and valid to derive this linear and additive model because the

correlation matrix as shown in Table 8 reveals that most of the top-10 weighted

KPIs are not highly correlated with each other at 1% significance level. In fact,

the linear and additive model has been widely applied in many different areas.

Frisch et al. (1992) developed the Quality of Life Inventory (QOLI), a measure of

life satisfaction, which was based on an empirically validated, linear, and

additive model. Nielsen (2001) introduced semi-parametric unmixing, which

was based on a linear and additive model with a non-linear smooth function to

represent end-member spectra. Ramaswami and Roe (2004) stated that a linear

additive model can be useful in any kind of risk analysis where it is vital to

distinguish between systemic and non-systemic risks. Yeung et al (2007 &

2009a) developed a Partnering Performance Index (PPI) for measuring the

partnering performance of construction projects in Hong Kong by using a linear

and additive model. Similarly, Yeung et al. (2009b), which was based on a

linear and additive model, developed a Performance Index (PI) for measuring

the performance of relationship-based construction projects in Australia. Xu et

al. (2010) developed a fuzzy risk allocation model for PPP projects in Mainland

China. By using a linear and additive model, the Risk Carrying Capability

Index (RCCI) was developed to evaluate the acceptable risk level of each project

participant. Supporting evidences concerning the appropriateness of applying

a linear and additive model in this research study were provided because

acceptable levels of model accuracy have been achieved in these studies (please



11

refer to the table in Appendix 1). Practically speaking, it is simpler and easier

to use this model to measure the performance of construction projects in the

Hong Kong construction industry (Nielsen 2001; Xu et al. 2010) because the

procedures involved in the calculations are much simpler and easier when

compared with a non-linear and multiplicative model.

In order to provide a greater flexibility to assessors to objectively, reliably and

practically use the model, it is more appropriate to define reasonable

quantitative ranges (QRs) for each KPI. In this research, a construction project

with “good” time performance would be one between 4.5 scores and 5.5 scores

(based on a 7-point Likert scale where 1 = Worst; 2 = Very Bad; 3 = Bad; 4 =

Average; 5 = Good; 6 = Very Good; and 7 = Excellent). In this example, the

lower boundary for the “good” time performance was simply taken as the

average value of the “average” time performance and “good” time performance,

whereas, the upper boundary for the “good” time performance was taken as the

average value of the “good” time performance and the “very good” time

performance. Similar approaches were adopted by Chow and Ng (2007) and

Yeung et al. (2008). It is recognized that applying fuzzy membership functions

to derive QRs for each KPI is an appropriate means for measurements involving

fuzziness. However, to achieve this, another empirical questionnaire survey

needs to be conducted which has not yet been executed in the current study

because of time and cost constraints. An equally useful, valid, and scientific

method as adopted by Chow and Ng (2007) in taking the average value of the

lower and upper limits of each two consecutive performance levels to derive the

QRs was adopted in the current study.

Case Studies – Application of KPIs and CPI

A set of KPIs and the CPI have been developed in the previous section. In

order to demonstrate the application of KPIs and CPI to measure the

performance of construction projects in Hong Kong, three case studies were

examined. The scope of analysis under each case study covers the project

performance of safety, cost, time, quality, client’s satisfaction, effectiveness of

communication, end user’s satisfaction, effectiveness of planning, functionality,

together with environmental performance. Table 9 shows the summary of the

background information and the results of different KPIs and CPI of these three

case studies in Hong Kong. The details of each case study are discussed in the

following subsections.



12

Case 1 – Civil Engineering Project

It is a civil engineering work, with total contract duration of 92 months and a

total contract sum of US$192.3077 million (HK$1,500 million). The project was

procured with sequential traditional method with tendering open to listed

contractors and the payment mechanism was lump sum with bills of

provisional quantities (re-measurement). The survey respondent perceived

that 8 out of 10 weighted KPIs were constructed with “excellent” performance.

The 8 weighted KPIs included: (1) safety performance; (2) time performance; (3)

quality performance; (4) client’s satisfaction; (5) effectiveness of communications;

(6) end user’s satisfaction; (7) functionality; and (8) environmental performance.

The other 2 weighted KPIs, including cost performance and effectiveness of

planning, were perceived to have “very good” performance. As a whole, a CPI

score of 6.80 out of a total of 7 was computed which suggests a project with

“excellent” performance.

Case 2 – Building Project

It is a building work, with total contract duration of 30 months and a total

contract sum of US$58.98 million (HK$460 million). The project was tendered

with restricted tendering method and the payment mechanism was lump sum

with bills of quantities. It was procured with management contracting. The

survey respondent perceived that 9 out of 10 weighted KPIs were constructed

with “good” performance. The 9 weighted KPIs were: (1) safety performance;

(2) cost performance; (3) time performance; (4) quality performance; (5) client’s

satisfaction; (6) effectiveness of communications; (7) end user’s satisfaction; (8)

functionality; and (9) environmental performance. The remaining weighted

KPI, effectiveness of planning, was perceived to have “average” performance.

As a whole, the CPI was found to be 4.90 (out of a total of 7), suggesting a

project with “good” performance.

Case 3 – Building Project

It is a building work, with total contract duration of 48 months and a total

contract sum of US$192.31 million (HK$1,500 million). The project was

procured with sequential traditional method with tendering open to listed

contractors and the payment mechanism was lump sum with bills of quantities.

The survey respondent perceived that 5 out of 10 weighted KPIs were

constructed with “very bad” performance. These included: (1) safety

performance; (2) cost performance; (3) time performance; (4) quality

performance; and (5) client’s satisfaction. Four of 10 weighted KPIs were



13

perceived to have “average” performance, which included: (1) effectiveness of

communications; (2) end user’s satisfaction; (3) functionality; and (4)

environmental performance. The remaining weighted KPI, effectiveness of

planning, was perceived to have “bad” performance. As a whole, the CPI

stood at 2.86 (out of a total of 7), implying a project with “bad” performance.

Significance and Limitations of the Research

The development of a series of both weighted leading and lagging KPIs is very

important. Otherwise, without an ex-ante evaluation of the performance level of

construction projects, project managers would find it difficult to allocate their

limited resources in the most efficient and effective way. Their decisions are

often made by perceived intuition, and they cannot ensure whether their actions

are correct or not. The proposed model provides an objective basis for not

only measuring but indeed predicting the performance of an on-going

construction project. The prediction of project performance may be conducted

on an on-going project once in a month (i.e. at the end of each month). The

model can be applied to measure, evaluate, and monitor the performance of

their individual construction project at different stages of the project life cycle

and right up to completion. The CPI not only enhances the understanding of

clients, contractors, and consultants to run a successful construction project, but

also it sets a solid base for industrial practitioners to measure, evaluate, and

monitor the performance of their individual construction projects right up to

completion. In addition, the CPI helps set a benchmark for measuring the

performance of a portfolio of construction projects within an organization, so

that this can help improve the overall performance of the firm through proper

monitoring and control of its projects. It also provides valuable insights into

developing a strong and comprehensive base for future research, for instance,

identifying critical success factors for current construction projects and then

developing a best practice framework for them. A similar research method can

also be extended to develop a benchmarking model for construction projects

internationally.

However, it is worth noting that it is likely for different evaluators to have their

own semantic interpretation of each KPI. For instance, an evaluator may use

“Percentage of conformance to the specifications” to measure quality

performance while another assessor may adopt “Number of non-conformance

reports generated per month” to measure it. Even though a mutually agreed

set of linguistic interpretations exists, its qualitative nature could lead to

subjective judgment instead of evidence-based consideration. Thus it is

desirable to identify appropriate quantitative interpretations/indicators (QIs) for



14

each KPI in order to avoid any possible discrepancies in interpreting the

meaning of each KPI and provide objective evaluation results based on

quantitative evidence. By incorporating these quantitative indicators into the

evaluation process, evaluators could perform their evaluation based on

quantitative evidence.

Conclusions

KPIs in construction have been discussed for many years, but there are not

many comprehensive and systematic studies on both leading and lagging KPIs.

This study has applied RIM to develop a model to objectively measure and

forecast the performance of construction projects in the Hong Kong

construction industry. The descending order of the top-10 weighted KPIs

identified were found to be: (1) safety performance, with the normalized

weighting of 0.1086; (2) cost performance, with the normalized weighting of

0.1058; (3) time performance, with the normalized weighting of 0.1058; (4)

quality performance, with the normalized weighting of 0.1018; (5) client’s

satisfaction, with the normalized weighting of 0.0991; (6) effectiveness of

communication, with the normalized value of 0.0969; (7) end user’s satisfaction,

with the normalized value of 0.0968; (8) effectiveness of planning, with the

normalized value of 0.0959; (9) functionality, with the normalized value of

0.0955; and (10) environmental performance, with the normalized value of

0.0937. The KPIs’ framework for construction projects helps develop a CPI and

sets a benchmark for measuring the performance of construction projects in

Hong Kong. Different construction projects can then be evaluated and

compared objectively based on this CPI established. As a result, construction

senior executives and project managers can use the CPI to measure, monitor,

and upgrade the performance of their construction projects at different stages of

the project life cycle. It also deepens the current body of knowledge and

understanding of both academics and practitioners in the construction industry

to achieve outstanding construction project performance.

Acknowledgement

The authors gratefully acknowledge the Department of Building and Real

Estate of The Hong Kong Polytechnic University for providing funding to

support this research study via the Departmental Postdoctoral Fellowship

Scheme (Project Account Code: BRE-1-ZV97).



15

References

Beatham, S., Anumba, C., Thorpe, T., and Hedges, I. (2004). “KPIs: a critical

appraisal of their use in construction.” Benchmarking: An International

Journal, 11(1), 93-117.

Chan, A.P.C., Wong, F.K.W., and Lam, P.T.I. (2006). “Assessing quality

relationships in public housing: an empirical study.” International Journal

of Quality and Reliability Management, 23(8), 909-927.

Chan, A.P.C., and Chan, A.P.L. (2004). “Key performance indicators for

measuring construction success.” Benchmarking: An International Journal,

11(2), 203-221.

Chan, A.P.C., Scott, D., and Lam, E.W.M. (2002). “Framework of success criteria

for design/build projects.” Journal of Management in Engineering, ASCE,

18(3), 120-128.

Chau, K.W. (1997). “The ranking of construction management journals.”

Construction Management and Economics, 15(4), 387-398.

Chow, L.K., and Ng, S.T. (2007). “Expectation of performance levels pertinent to

consultant performance evaluation.” International Journal of Project

Management, 25(1), 90-103.

Costa, D.B., Formoso, C.T., Kagioglou, M., Alarcón, L.F., and Caldas, C.H. (2006).

“Benchmarking initiatives in the construction industry: lessons learned

and improvement opportunities.” Journal of Management in Engineering,

ASCE, 22(4), 158-167.

Cox, R.F., Issa, R.R.A., and Ahrens, D. (2003). “Management’s perception of key

performance indicators for construction.” Journal of Construction

Engineering and Management, ASCE, 129(2), 142-151.

Curtin, R.T. (1982). “Indicators of consumer behavior: The University of

Michigan Surveys of Consumers.” Public Opinion Quarterly, 46(3),

340-352.

El-Mashaleh, M.S., Minchin, R.E., and O’Brien, W.J. (2007). “Management of

construction firm performance using benchmarking.” Journal of

Management in Engineering, ASCE, 23(1), 10-17.

Frisch, M.B., Cornell, J., and Villanueva, M. (1992). “Clinical validation of the

quality of life inventory: a measure of life satisfaction for use in treatment

planning and outcome assessment.” Psychological Assessment, 4(1), 92-101.

Gammon Construction. (2010).

(http://www.gammonconstruction.com/html/about-us/values.html) (May

2010).

Hong Kong Labor Department (2009).Occupational safety and health statistics

bulletin Issue No. 7 2009. Hong Kong Special Administrative Region:

Government Printer, 2009.



16

Ke, Y.J., Wang, S.Q., Chan, A.P.C., and Cheung, E. (2009). “Research trend of

Public-Private Partnership (PPP) in construction journals.” Journal of

Construction Engineering and Management, ASCE, 135(10), 1076-1086.

Lam E.W.M., Chan, A.P.C., and Chan, D.W.M. (2007). “Benchmarking the

performance of design-build projects: development of project success

index.” Benchmarking: An International Journal, 14(5), 624-638.

Lam E.W.M., Chan, A.P.C., and Chan, D.W.M. (2008). “Determinants of

successful design-build projects.” Journal of Construction Engineering and

Management, ASCE, 134(5), 333-341.

Lee, S.H., Thomas, S.R., and Tucker, R.L. (2005). “Web-based benchmarking

system for the construction industry.” Journal of Construction Engineering

and Management, ASCE, 131(7), 790-798.

Ling, F.Y.Y., Low, S.P., Wang, S.Q., and Egbelakin, T. (2008). “Models for

predicting project performance in China using project management

practices adopted by foreign AEC firms.” Journal of Construction


Lo, S.M., Hu, B.Q., Liu, M., and Yuen, K.K. (2005). “On the use of reliability

interval method and grey relational model for fire safety ranking of

existing buildings.” Fire Technology, 41(4), 255-270.

Lo, S.M., Lu, J.A., Hu, Y.Q., and Fang, Z. (2001). “Incorporating reliability and

variance into weighting function of fire risk assessment for high-rise

buildings.” China Safety Science Journal, 11(5), 11-13 (Chinese).

Luu, V.T., Kim, S.Y., and Huynh, T.A. (2008). “Improving project management

performance of large contractors using benchmarking approach.”

International Journal of Project Management, 26(7), 758-769.

Moore, R.E. (1979). “Methods and applications of interval analysis.” Philadelphia:

SIAM.

Nielsen, A.A. (2001). “Spectral mixture analysis: linear and semi-parametric full

and iterated partial unmixing in multi-and hyperspectral image data.”

Journal of Mathematical Imaging and Vision, 15(1-2), 17-37.

Ramaswami, B., and Roe, T.L. (2004). “Aggregation in area-yield crop insurance:

the linear additive model.” American Journal of Agricultural Economics,

86(2), 420-431.

Ramirez, R.R, Alarcón, L.F.C., and Knights, P. (2004). “Benchmarking system for

evaluating management practices in the construction industry.” Journal of

Management in Engineering, ASCE, 20(3), 110-117.

Satty, T.L. (1980). “The analytic hierarchy process.” New York: McGraw Hill.

Saaty, T.L. (2008). “Decision making with the analytic hierarchy process.”

International Journal of Services Sciences, 1(1), 83-98.



17

Shields, J., and Silcock, G. (1986). “An application of the hierarchical approach to

fire safety.” Fire Safety Journal, 11(3), 235-242.

Westerveld, E. (2003). “The project excellence model: linking success criteria and

critical success factors.” International Journal of Project Management, 21(6),

411-418.

Xu, Y.L., Chan, A.P.C., and Yeung, J.F.Y. (2010). “Developing a fuzzy risk

allocation model for PPP projects in China.” Journal of Construction


Yeung, J.F.Y., Chan, A.P.C., and Chan, D.W.M. (2008). “Establishing quantitative

indicators for measuring the partnering performance of construction

projects in Hong Kong.” Construction Management and Economics, 26(3),

277-301.

Yeung, J.F.Y., Chan, A.P.C., and Chan, D.W.M. (2009a). “A computerized model

for measuring and benchmarking the partnering performance of

construction projects.” Automation in Construction, 18(8), 1099-1113.

Yeung, J.F.Y., Chan, A.P.C., and Chan, D.W.M. (2009b). “Developing a

performance index for relationship-based construction projects in

Australia: Delphi study.” Journal of Management in Engineering, ASCE,

25(2), 59-68.

Yeung, J.F.Y., Chan, A.P.C., Chan, D.W.M., and Li, L.K. (2007). “Development

of a partnering performance index (PPI) for construction projects in

Hong Kong: a Delphi study.” Construction Management and Economics,

25(12), 1219-1237.

Yiu, C.Y., Ho, H.K., Lo, S.M., Hu, B.Q. (2005). “Performance evaluation for cost

estimators by reliability interval method.” Journal of Construction




18

Table 1. Level of Relevance to Benchmarking Related Studies in Construction for Reviewed Journal Abstracts

Level of relevance to benchmarking studies in construction

Descriptor Total number of journal abstracts reviewed

Not relevant = 0 The aim of the paper is not relevant to benchmarking study in construction. The keywords selected for the Scopus database search are rarely mentioned in the abstract. In addition, the paper title and keywords do not reflect any relevance on benchmarking study.

379

Least relevant = 1 The aim of the paper is least relevant to benchmarking study in construction. The keywords selected for the Scopus database search are seldom mentioned in the abstract. Besides, the paper title and keywords only reflect a low level of relevance on benchmarking study.

70

Slightly relevant = 2 The aim of the paper is slightly relevant to benchmarking study in construction. The keywords selected for the Scopus database search are sometimes mentioned in the abstract. Moreover, the paper title and keywords reflect a moderate level of relevance on benchmarking study.

30

Relevant = 3 The aim of the paper is relevant to benchmarking study in construction. The keywords selected for the Scopus database search are mentioned in the abstract. In addition, the paper title and keywords reflect a high level of relevance on benchmarking study.

79

Very relevant = 4 The aim of the paper is very relevant to benchmarking study in construction. The keywords selected for the Scopus database search are frequently mentioned in the abstract. In addition, the paper title and keywords reflect a very high level of relevance on benchmarking study.

12

Most relevant = 5 The aim of the paper is extremely relevant to benchmarking study in construction. The keywords selected for the Scopus database search are very frequently mentioned in the abstract. In addition, the paper title and keywords reflect an extremely high level of relevance on benchmarking study.

2

Total

572



19

Table 2. Performance Indicators for Construction Projects Performance Indicators

Chan

et

al.

(200

2)

Cox e

t al

. (2

003

)

Wes

terv

eld

(2003

)

Bea

tham

et

al. (2

004

)

Chan

and

Chan

(20

04)

Ram

irez

et

al. (2

00

4)

Lee

et

al. (2

00

5)

Cost

a et

al.

(200

6)

El-

Mas

hal

eh e

t al

. (2

007)

Yeu

ng

et

al. (2

007

&

2009

)

Lam

et

al. (2

00

8)

Lin

g e

t al

. (2

008)

Luu e

t al

. (2

008

)

To

tal

1. Schedule/Time performance √ √ √ √ √ √ √ √ √ √ 10 2. Cost performance √ √ √ √ √ √ √ √ √ √ 10 3. Safety performance/Accident rate √ √ √ √ √ √ √ √ √ 9 4. Profit/Profitability √ √ √ √ √ √ √ 7 5. Client’s satisfaction √ √ √ √ √ √ √ 7 6. Quality performance √ √ √ √ √ √ 6 7. Total field rework factor (leading measure)/Scope of rework √ √ √ √ √ 5 8. Productivity performance (leading measure) √ √ √ √ √ 5 9. Environmental performance √ √ √ √ 4 10. End user’s satisfaction (on services or on products or on projects) √ √ √ √ 4 11. Project cost growth (owner) √ √ √ 3 12. Project budget factor (contractor) √ √ √ 3 13. Functionality √ √ √ 3 14. Recordable incident rate (contractor) √ √ √ 3 15. Project schedule factor √ √ 2 16. Project schedule growth √ √ 2 17. Lost workday case incident rate √ √ 2 18. Change cost factor √ √ 2 19. Defects √ √ 2 20. Predictability cost √ √ 2 21. Predictability time √ √ 2 22. Extent and layer of subcontracting √ √ 2 23. Efficiency of direct labor √ √ 2 24. Effectiveness of risk management (leading measure) √ √ 2 25. Effectiveness of planning (leading measure) √ √ 2 26. Project team satisfaction (leading measure) √ √ 2 27. Provision of training courses (leading measure) √ √ 2 28. Effectiveness of material management and resource management (leading measure) √ √ 2 29. Professional image establishment √ √ 2 30. Innovation and improvement √ √ 2 31. Claim avoidance √ √ 2 32. Effectiveness of communication (leading measure) √ √ 2 33. Staff turnover √ √ 2 34. Experience modification rating (EMR) √ 1 35. Project management expenses √ 1 36. Total project duration √ 1 37. Factor phase cost growth (owner data only) (leading measure) √ 1 38. Phase duration factor (leading measure) √ 1 39. Construction phase duration (leading measure) √ 1 40. Percentage of plan completed (leading measure) √ 1 41. Supplier performance (leading measure) √ 1 42. No. of nonconformity in audits (leading measure) √ 1 43. Rate of employees trained (leading measure) √ 1 44. Quality management system (QMS) √ 1 45. Project team performance √ 1



20

Table 2. (Continued) 46. Change management √ 1 47. Labor safety management √ 1 48. Phase cost factor (owner) √ 1 49. Zero accident techniques practice use (owner) √ 1 50. Litigation avoidance √ 1 51. Dispute avoidance √ 1 52. Harmonious working relationships √ 1 53. Long-term business relationship √ 1 54. Top management support √ 1 55. Employee’s attitude √ 1 56. Understanding client needs √ 1

Total 8 6 4 15 9 10 5 30 15 20 7 5 8



21

Table 3. Performance Indicators after Consolidation for Construction Projects Performance indicators

Ch

an e

t al

. (2

002

)

Co

x e

t al

. (2

003)

Wes

terv

eld

(2

003)

Bea

tham

et

al. (2

004)

Ch

an a

nd

Chan

(200

4)

Ram

irez

et

al. (2

004

)

Lee

et

al. (2

005

)

Co

sta

et a

l. (

200

6)

El-

Mas

hal

eh e

t al

. (2

007)

Yeu

ng e

t al

. (2

007

& 2

00

9)

Lam

et

al. (2

00

8)

Lin

g e

t al

. (2

008)

Lu

u e

t al

. (2

008)

To

tal

1. Cost performance √ √ √ √ √ √ √ √ √ √ √ √ √ 13

2. Schedule/Time performance √ √ √ √ √ √ √ √ √ √ √ √ 12

3. Safety performance/Accident rate √ √ √ √ √ √ √ √ √ √ √ 11

4. Quality performance √ √ √ √ √ √ √ √ √ √ √ 11

5. Profit/Profitability √ √ √ √ √ √ √ 7

6. Client’s satisfaction √ √ √ √ √ √ √ 7

7. Productivity performance (leading measure) √ √ √ √ √ 5

8. Environmental performance √ √ √ √ 4

9. End user’s satisfaction (on services or on products or on projects) √ √ √ √ 4

10. Functionality √ √ √ 3

11. Project team satisfaction (leading measure) √ √ √ 3

12. Extent and layer of subcontracting √ √ 2

13. Effectiveness of risk management (leading measure) √ √ 2

14. Effectiveness of planning (leading measure) √ √ 2

15. Provision of training courses (leading measure) √ √ 2

16. Effectiveness of material management and resource management (leading measure) √ √ 2

17. Professional image establishment √ √ 2

18. Innovation and improvement √ √ 2

19. Claim avoidance √ √ 2

20. Effectiveness of communication (leading measure) √ √ 2

21. Project management expenses √ 1

22. Supplier performance (leading measure) √ 1

23. Change management √ 1

24. Harmonious working relationships √ 1



22

Table 3. (Continued) 25. Long-term business relationship √ 1

26. Top management support √ 1

27. Employee’s attitude √ 1

28. Understanding client needs √ 1 Total 8 5 4 14 9 7 2 13 7 17 7 5 8



23

Table 4. Type of Working Organization for Survey Respondents

Type of organization Frequency Valid percent Cumulative percent

Client organization 143 61.4 61.4

Main contractor 39 16.7 78.1

Architectural consultant 8 3.4 81.5

Engineering consultant 13 5.6 87.1

Quantity Surveying consultant 4 1.7 88.8

Project management consultant 5 2.1 91.0

Subcontractor 7 3.0 94.0

Supplier/Manufacturer 2 .9 94.8

Others 12 5.2 100.0

Total 233 100.0

Table 5. Years of Professional Working Experience for Survey Respondents

Years of professional working experience

Frequency Valid percent Cumulative percent

Below 5 years 8 3.4 3.4

5-10 years 21 9.0 12.4

11-15 years 30 12.9 25.3

16-20 years 33 14.2 39.5

Over 20 years 141 60.5 100.0

Total 233 100.0



24

Table 6. Average Interval Grades for 20 KPIs for Construction Projects in Hong Kong

KPIs for construction projects in Hong Kong Average interval grades 1. Time performance [3.174, 4.680] 2. Cost performance [3.740, 4.656] 3. Safety performance [3.916, 4.700] 4. Profit/Profitability [2.987, 3.923] 5. Client’s satisfaction [3.479, 4.382] 6. Quality performance [3.561, 4.512] 7. Productivity performance [3.054, 3.976] 8. Environmental performance [3.251, 4.179] 9. End user’s satisfaction [3.394, 4.284] 10. Functionality [3.324, 4.255] 11. Extent and layer of subcontracting [2.528, 3.465] 12. Effectiveness of risk management [3.119, 4.031] 13. Effectiveness of planning [3.341, 4.273] 14. Project team satisfaction [2.866, 3.833] 15. Provision of training courses [2.439, 3.395] 16. Effectiveness of material and resource management [2.958, 3.879] 17. Professional image establishment [2.907, 3.799] 18. Innovation and improvement [2.788, 3.709] 19. Dispute avoidance [2.867, 3.835] 20. Effectiveness of communication [3.395, 4.288]

Table 7. Results of Reliability Interval Method (RIM) for KPIs Related to Construction Projects in Hong Kong

KPIs for construction projects in Hong Kong Weighting Reliability Centre variance (CV)

Interval variance (IV)

1. Safety performance 0.05914 0.81974 0.32490 0.85009 2. Cost performance 0.05763 0.92704 0.28111 0.93499 3. Time performance 0.05762 0.95279 0.24727 0.91820 4. Quality performance 0.05541 0.92704 0.30568 1.02566 5. Client’s satisfaction 0.05396 0.87983 0.39514 1.12176 6. Effectiveness of communication 0.05274 0.85837 0.47722 1.23273 7. End user’s satisfaction 0.05270 0.85408 0.42093 1.14217 8. Effectiveness of planning 0.05226 0.87554 0.39412 1.14699 9. Functionality 0.05202 0.90558 0.36175 1.08857 10. Environmental performance 0.05100 0.84979 0.47001 1.20828 11. Effectiveness of risk management 0.04908 0.75966 0.51977 1.24782 12. Productivity performance 0.04826 0.84120 0.38984 0.99513 13. Profit/Profitability 0.04744 0.59307 0.85748 1.71176 14. Effectiveness of material and resource

management 0.04693 0.77682 0.47352 1.17108

15. Professional image establishment 0.04603 0.68240 0.73076 1.56802 16. Dispute avoidance 0.04601 0.75966 0.57985 1.40981 17. Project team satisfaction 0.04599 0.76395 0.56225 1.40423 18. Innovation and improvement 0.04460 0.74678 0.62209 1.42950 19. Extent and layer of subcontracting 0.04114 0.75862 0.66686 1.53460 20. Provision of training courses 0.04005 0.77682 0.61361 1.51269



25

Table 8. Correlation Matrix among the Top Ten Weighted Leading and Lagging KPIs (Sample Size = 278)

Correlation matrix

Safety performance

Cost performance

Time performance

Quality performance

Client’s satisfaction

Effectiveness of communication

End user’s satisfaction

Effectiveness of planning

Functionality

Environmental performance

Safety performance

1 0.414** 0.402** 0.504** 0.506** 0.468** 0.351** 0.345** 0.385** 0.444**

Cost performance

1 0.495** 0.488** 0.573** 0.461** 0.412** 0.456** 0.426** 0.359**

Time performance

1 0.513** 0.612** 0.482** 0.491** 0.538** 0.464** 0.408**

Quality performance

1 0.749** 0.656** 0.689** 0.526** 0.616** 0.540**

Client’s satisfaction

1 0.631** 0.687** 0.602** 0.589** 0.541**

Effectiveness of communication

1 0.629** 0.533** 0.560** 0.491**

End user’s satisfaction

1 0.552** 0.649** 0.457**

Effectiveness of planning

1 0.588** 0.504**

Functionality

1 0.618**

Environmental performance

1

**. Correlation is significant at the 0.01 level (two-tailed)



26

Table 9. Case Studies: Application of KPIs and CPI

Case study Case 1 Case 2 Case 3 Background Nature of project Civil work Building work Building work Procurement method Sequential traditional Management contracting Sequential traditional Tendering method Tendering open to listed contractors Restricted tendering Tendering open to listed

contractors Payment mechanism Lump sum with bills of provisional

quantities (re-measurement) Lump sum with bills of

quantities Lump sum with bills of

quantities Total contract duration 92 months 30 months 48 months Total contract sum US$192.31 million

(HK$1500 million) US$58.98 million (HK$460 million)

US$192.31 million (HK$1500 million)

KPIs survey result Safety Performance Excellent

(7 out of 7 scores) Good

(5 out of 7 scores) Very Bad

(2 out of 7 scores) Cost Performance Very Good



(2 out of 7 scores) Time Performance Excellent



(2 out of 7 scores) Quality Performance Excellent



(2 out of 7 scores) Client’s Satisfaction Excellent



(2 out of 7 scores) Effectiveness of Communications

Excellent (7 out of 7 scores)

Good (5 out of 7 scores)

Average (4 out of 7 scores)

End User’s Satisfaction Excellent (7 out of 7 scores)



Effectiveness of Planning Very Good (6 out of 7 scores)


Bad (3 out of 7 scores)

Functionality Excellent (7 out of 7 scores)



Environmental Performance

Excellent (7 out of 7 scores)



Composite Performance Index (CPI)

6.7976 (out of a total of 7) 4.9034 (out of a total of 7) 2.8615 (out of a total of 7)



27

Appendix. Comparisons of Levels of Model Accuracy between the Results of Former

Studies and the Current Study

Study Model

1. Frisch et al. (1992) Frisch et al. (1992) developed the Quality of Life Inventory (QOLI), a measure of life satisfaction, which

was based on an empirically validated, linear, and additive model. Test-retest coefficients for the QQLI

ranged from 0.80 to 0.91, and internal consistency coefficients ranged from 0.77 to 0.89 across 3 clinical

and 3 non-clinical samples.

2. Nielsen (2001) Nielsen (2001) introduced semi-parametric unmixing, which was based on a linear and additive model

with a non-linear smooth function to represent end-member spectra. An example with two generated

bands shows that both full unmixing, the Constrained Energy Minimization (CIM), the Iterated CEM and

Target Constrained Interference Minimized Filter (TCIMF) methods perform well. A case study with a

30 bands subset of AVIRIS data shows the utility of full unmixing, Spectral Angle Mapping (SAM), CEM

and Iterated CEM to more realistic data.

3. Ramaswami and

Roe (2004)

Ramaswami and Roe (2004) stated that a linear additive model can be useful in any kind of risk analysis

where it is vital to distinguish between systemic and non-systemic risks. The authors have

mathematically derived the linear additive model (LAM) from aggregation of micro-production functions.

4. Yeung et al. (2007,

2009a).

Yeung et al (2007 and 2009a) developed a Partnering Performance Index (PPI) for measuring the

partnering performance of construction projects in Hong Kong by using a linear and additive model.

Three case studies were used to validate the usefulness of the model. In fact, the model has also been

validated by seven independent experts who had not previously been involved in the Delphi questionnaire

surveys for proper validation.

5. Yeung et al.

(2009b)

Yeung et al. (2009b), based on a linear and additive model, developed a Performance Index (PI) for

measuring the performance of relationship-based construction projects in Australia. Three case studies

were used to validate the usefulness of the model. In addition, it was noted that the Delphi survey

technique by its inherent nature serves as a self-validating mechanism because each expert was given a

chance to re-evaluate their scores with reference to the consolidated mean scores as assessed by other

experts.

6. Xu et al. (2010)

Xu et al. (2010) developed a fuzzy risk allocation model for PPP projects in Mainland China. By using a

linear and additive model, the Risk Carrying Capability Index (RCCI) was developed to evaluate the

acceptable risk level of each project participant. An illustrative example was used to demonstrate the

usefulness of the model.

7. Current study

Yeung et al. (2013), based on a linear and additive model, developed a benchmarking model for

construction project in Hong Kong by incorporating both leading and lagging KPIs and applying the

Reliability Interval Method (RIM). Three case studies were used to validate the usefulness of the model.

developing a benchmarking model for construction...

Documents